How data, data science, and asking the right questions can help us tackle an information crisis

By Carlotta Alfonsi, Fiona Cece, and Craig Matasick

People and governments alike continue to struggle with responding to the challenges of the COVID-19 pandemic. The real-world health, economic and social challenges faced, however, are exacerbated by the rampant mis- and disinformation around COVID-19 that has spread rapidly since the pandemic began. The so-called “infodemic” poses a great risk to the health and safety of society – as well as the response to it (globally, regionally, nationally, and locally). For example, the wide and rapid spread of mis/disinformation can reduce the effectiveness of policy implementation or observance of public health rules, increase vaccine hesitancy, polarize public debate, among other negative outcomes.

To combat the infodemic, government and non-government actors across various sectors will need to work together to help develop a whole-of-society response to the challenges we face. One example of a multi-sectoral collaborative initiative is that by the French Ambassador for Digital Rights, which includes a free Twitter API that can be easily deployed by citizens to identify clusters of bots on the social media platform.

But the challenges of mis- and disinformation did not start – nor will they stop – with the Coronavirus pandemic. The public’s confidence in information has been undermined and the ability for citizens to agree on a set of facts is being fractionalized.

The consequences of mis/disinformation permeate all aspects of the policy discussion and pose a risk to democratic governance. Indeed, information disorders can damage the integrity of electoral processes, hinder dialogue between the government and the public, and break down societal cohesion.

Stakeholders from different sectors have been pooling diverse expertise and are collaborating to tackle this complex phenomenon. However, a common agenda is needed to channel these efforts for greater impact.

The New Science of Questioning: Are we asking the right questions?

One initiative working toward this end is the 100 Questions Initiative by The Governance Lab (The GovLab) at NYU Tandon School of Engineering. In partnership with the Organization for Economic Cooperation and Development (OECD), the initiative has drawn on the expertise of over 100 “bilinguals” – experts in both data science and disinformation (including the Global Disinformation Index’s co-founders Clare Melford and Danny Rogers) – to identify the 10 most pressing questions related to disinformation that can be addressed using data and data science.

The Initiative leverages a participatory methodology to map out these questions with the help of the bilinguals. Over the course of the past few months, participants have developed, curated, and prioritized the questions.

The following ten questions are the result of this process (more details for each question is available at the Disinformation hub on the 100 Questions website):

Top 10 Disinformation Questions:

1. Are there specific factors that make people more or less likely to consume and spread disinformation? Do those factors help us understand how people or groups can access and believe corrective information?
2. How does disinformation contribute to online hate speech, radicalization and real-life violence? What are the factors that heighten the risk of disinformation resulting in hate speech, harassment, or violence? How do narratives target particular groups or lead to their silencing?
3. How can we monitor and address the spread of mis/disinformation that happens in closed messaging platforms?
4. What are the effects of disinformation on political opinion or behavior? How can these effects be measured, and at what point do these effects threaten democratic governance?
5. How can we measure the different sorts of effects of exposure to mis and disinformation and how can we identify the conditions under which mis/disinformation causes actual or potential harmful effects?
6. What are the most pressing concerns related to mis/disinformation for vulnerable and marginalized populations? How can their experiences be better understood, and how do their experiences inform their reaction to misinformation?
7. What interventions are most effective in combating disinformation, including but not limited to preventative measures, regulatory responses, and citizen engagement (distinct from other algorithmic, corporate, and governmental responses)? How do we think proactively about how to prepare for mis/disinformation targeted at populations who are not yet online/mobile yet?
8. What types of cost and incentive structures could be applied to platforms to prevent them from profiting off mis/disinformation? And how can we create a business model that rewards platforms for reducing disinformation on their platforms?
9. What is the role of traditional media in amplifying disinformation? How have changes in traditional media markets globally affected the spread of disinformation?
10. What are the features and characteristics of disinformation campaigns that go viral or spread to other media? To what extent are these different from viral posts in general? Why do people believe mis/disinformation? What is the tipping point when a peripheral conspiracy theory becomes mainstream disinformation which makes a measurable social impact?

We now turn to you, the public, to hear your voice and ideas.

To vote on which questions are most important to you, please visit disinformation.the100questions.org, and make your voice heard!

The ultimate goal is to establish one or more data collaboratives that can generate answers to the questions at hand. (Data collaboratives are flexible structures that allow pooling of data and expertise across sectors, often resulting in new insights and public sector innovations).

This guest blog was written and provided by The 100 Questions Initiative, at The Governance Lab at NYU.